Research And Implementation Of Tunnel Face Image Processing System

Image is the basic information in modern information society. Digital image processing technology is a useful technology which has great economic and social benefit, and be widely used in many fields.This project's research is based on the ground of the tunnel development. In the processing of tunnel excavation, we got a series of tunnel face images by digital photographic equipment. These images will be analyzed and processed by using digital image processing and pattern recognition technology. Then, we can acquire all terrene edge information and fulfill tunnel 3-dimension rebuilding and forecasting at last. Compared with traditional analysis methods, the digital image processing technology used in tunnel excavation is more efficient, which can largely save time of tunnel face image analysis. Meanwhile, it reduces the cost. Therefore its applications are very extensive.First of all, the paper introduces the project's research background and major research content, analyses what problems will be resolved in the tunnel face image processing system, and what are emphases and difficulties in the processing of system's realization. Then, this paper simply introduces the digital image processing technology and oriented object programming design technology used in system's realization. This paper concerns the tunnel face image processing system. Digital image processing technology is the core of the system. This paper amply studies the possible arithmetic used in different parts, and gives the processing result as well. In the image fore processing, this paper mainly concerns about image geometry emendation by polynomial equation, image gray-scale linear transform, image gray-scale window transform and image spatial filter processing and so on. In image edge detection, this paper mainly studies how to detect tunnel face image's edge through Canny operator. In image segmentation, this paper's emphasis is how to segment image through growth algorithm based on area. Because of tunnel face image's complexity, it is very difficult to segment image automatically. Therefore, system provides a method called "Intelligent scissor" to segment image semiautomatic or manually. It can segment image more accurately and satisfy all kinds of demand. This paper's final purpose is to build a tunnel 3-dimension model and forecast tunnel structure. Therefore, we need to acquire object character parameters and apply pattern recognition. Then, all relevant objects among the adjoining image slices will be interlinked accurately.The tunnel face image processing system based on this paper has friendly user operation interface. The result shows that the graph and image processing technology and the pattern recognition given in the paper can initially fulfill the anticipative achievements and provide essential data and help to subsequent tunnel 3-dimension rebuilding and forecasting.